Title:	A TRAINING FILE RECOMMENDATION SYSTEM FOR THE LEARNER OF FEATURE-BASED DESIGN SYSTEMS
Author(s):	Li-chieh Chen , Yen-fu Chen
ISBN:	972-8924-16-X
Editors:	Pedro Isaías, Maggie McPherson and Frank Bannister
Year:	2006
Edition:	2
Keywords:	Recommendation system, Training file, Personalization, Content-based filtering, Collaborative filtering.
Type:	Short Paper
First Page:	39
Last Page:	43
Language:	English
Cover:
Full Contents:	click to dowload
Paper Abstract:	Training files are important for self-learning while using a feature-based design system. However, selecting appropriate training files that are suitable for a learner with specific knowledge and skill levels is a difficult task. Therefore, in order to facilitate self-learning, the objective of this research is to develop a personalized recommendation system for the training files of a feature-based design system. The authors employ content-based filtering and collaborative filtering to construct the system. In content-based filtering, a vector-based representation scheme is used to model the training file and learner profile. The keywords for searching training files are recorded in the database with their important ratings. The similarity is then quantified by the cosine of the angle between these two vectors and converted into the prediction of rating. In collaborative filtering, usefulness ratings for files are provided by the learners and recorded in the database. These ratings are used to predict the ratings for new items for specific learners. The predicted rating of each file is then calculated by combining the results from content-based and collaborative filtering linearly. Based on the predicted rating, training files that are useful to a learner are retrieved and listed in a descending order in the user interface of this system. The result shows that learners can obtain appropriate training files effectively.
	Go Back